Automatic Detection of Abnormal Regions in Endoscopic Images of Esophagus Using a Skin Color Model

 P.S. Hiremath and Humnabad Iranna Y.

Department of P.G. Studies and Research in Computer Science, Gulbarga University, Gulbarga, Karnataka, India

Abstract:

An image segmentation system is proposed for the detection of abnormal regions (esophageal cancer) of esophagus in endoscopic color image based on skin-color model. The illumination sensitivity influences the detection of abnormal regions of esophagus in endoscopic images. We have used nonlinearly transformed YC_bC_r color space devoid of luma-dependency for detection of abnormal regions in endoscopic images. The proposed method is compared with 3d interval method and the piecewise linear skin-color model. The experimental results demonstrate the efficiency of the proposed method for model-based diagnosis of abnormal regions in esophagus endoscopic images.

Keywords: Endoscopy, Esophagus, abnormal region, cancer.

( Full Paper 1 MB)

BibTex:

@ARTICLE{P1120647001,

AUTHOR = {P.S. Hiremath and Humnabad Iranna Y.},

TITLE = {Automatic Detection of Abnormal Regions in Endoscopic Images of Esophagus Using a Skin Color Model},

JOURNAL ={ICGST  International Journal on Artificial Intelligence and Machine Learning, AIML},

}

( Full Paper 1 MB)